通过动态调整生成过程的精度，让图像生成更快更清晰。
TMPDiff: Temporal Mixed-Precision for Diffusion Models
- 按不同步骤分配不同精度，避免全程低精度带来的损失。
- 在4个主流模型上实现2.5倍加速，仍保持90%图像质量。
- 适合追求高速与高质量平衡的生成系统部署者。
扩散模型是文本生成图像的主流方法，但其迭代去噪过程导致推理延迟高。量化通过使用更低比特数减少计算时间，但通常在整个去噪步骤中采用固定精度，未充分利用优化空间。本文提出TMPDiff，一种针对扩散模型的时间混合精度框架，为不同去噪步骤分配不同数值精度。我们假设量化误差随步骤累积，经实验验证成立。基于此，设计了一种基于二分搜索的自适应算法，以线性复杂度完成每步精度分配，将原本指数级的搜索问题简化。在四个顶尖扩散模型和三个数据集上，TMPDiff在相同加速比下均优于统一精度基线，感知质量提升10%至20%。在FLUX.1-dev上，以16位推理2.5倍加速，仍达到全精度模型90%的SSIM。
原文摘要 · Abstract (English)
Diffusion models are the go-to method for Text-to-Image generation, but their iterative denoising processes has high inference latency. Quantization reduces compute time by using lower bitwidths, but applies a fixed precision across all denoising timesteps, leaving an entire optimization axis unexplored. We propose TMPDiff, a temporal mixed-precision framework for diffusion models that assigns different numeric precision to different denoising timesteps. We hypothesize that quantization errors accumulate additively across timesteps, which we then validate experimentally. Based on our observations, we develop an adaptive bisectioning-based algorithm, which assigns per-step precisions with linear evaluation complexity, reducing an otherwise exponential search problem. Across four state-of-the-art diffusion models and three datasets, TMPDiff consistently outperforms uniform-precision baselines at matched speedup, achieving 10 to 20% improvement in perceptual quality. On FLUX.1-dev, TMPDiff achieves 90% SSIM relative to the full-precision model at a speedup of 2.5x over 16-bit inference.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。